]> jfr.im git - yt-dlp.git/blobdiff - yt_dlp/extractor/youtube.py
[extractor] Better error message for DRM (#729)
[yt-dlp.git] / yt_dlp / extractor / youtube.py
index 365c2b95b960e20dbbc08cf6c347a261637db451..375eca8f8499c8cabae60fafbe4caeaebaf51b0b 100644 (file)
@@ -38,6 +38,7 @@
     format_field,
     int_or_none,
     intlist_to_bytes,
+    is_html,
     mimetype2ext,
     network_exceptions,
     orderedSet,
@@ -45,6 +46,7 @@
     parse_count,
     parse_duration,
     parse_iso8601,
+    parse_qs,
     qualities,
     remove_start,
     smuggle_url,
 )
 
 
-def parse_qs(url):
-    return compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
-
-
 # any clients starting with _ cannot be explicity requested by the user
 INNERTUBE_CLIENTS = {
     'web': {
@@ -100,6 +98,16 @@ def parse_qs(url):
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 67,
     },
+    'web_creator': {
+        'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8',
+        'INNERTUBE_CONTEXT': {
+            'client': {
+                'clientName': 'WEB_CREATOR',
+                'clientVersion': '1.20210621.00.00',
+            }
+        },
+        'INNERTUBE_CONTEXT_CLIENT_NAME': 62,
+    },
     'android': {
         'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8',
         'INNERTUBE_CONTEXT': {
@@ -131,6 +139,17 @@ def parse_qs(url):
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 21,
     },
+    'android_creator': {
+        'INNERTUBE_CONTEXT': {
+            'client': {
+                'clientName': 'ANDROID_CREATOR',
+                'clientVersion': '21.24.100',
+            },
+        },
+        'INNERTUBE_CONTEXT_CLIENT_NAME': 14
+    },
+    # ios has HLS live streams
+    # See: https://github.com/TeamNewPipe/NewPipeExtractor/issues/680
     'ios': {
         'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8',
         'INNERTUBE_CONTEXT': {
@@ -162,6 +181,17 @@ def parse_qs(url):
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 26
     },
+    'ios_creator': {
+        'INNERTUBE_CONTEXT': {
+            'client': {
+                'clientName': 'IOS_CREATOR',
+                'clientVersion': '21.24.100',
+            },
+        },
+        'INNERTUBE_CONTEXT_CLIENT_NAME': 15
+    },
+    # mweb has 'ultralow' formats
+    # See: https://github.com/yt-dlp/yt-dlp/pull/557
     'mweb': {
         'INNERTUBE_API_KEY': 'AIzaSyDCU8hByM-4DrUqRUYnGn-3llEO78bcxq8',
         'INNERTUBE_CONTEXT': {
@@ -176,11 +206,14 @@ def parse_qs(url):
 
 
 def build_innertube_clients():
+    third_party = {
+        'embedUrl': 'https://google.com',  # Can be any valid URL
+    }
     base_clients = ('android', 'web', 'ios', 'mweb')
     priority = qualities(base_clients[::-1])
 
     for client, ytcfg in tuple(INNERTUBE_CLIENTS.items()):
-        ytcfg.setdefault('INNERTUBE_API_KEY', 'AIzaSyDCU8hByM4DrUqRUYnGn3llEO78bcxq8')
+        ytcfg.setdefault('INNERTUBE_API_KEY', 'AIzaSyDCU8hByM-4DrUqRUYnGn-3llEO78bcxq8')
         ytcfg.setdefault('INNERTUBE_HOST', 'www.youtube.com')
         ytcfg['INNERTUBE_CONTEXT']['client'].setdefault('hl', 'en')
         ytcfg['priority'] = 10 * priority(client.split('_', 1)[0])
@@ -188,8 +221,10 @@ def build_innertube_clients():
         if client in base_clients:
             INNERTUBE_CLIENTS[f'{client}_agegate'] = agegate_ytcfg = copy.deepcopy(ytcfg)
             agegate_ytcfg['INNERTUBE_CONTEXT']['client']['clientScreen'] = 'EMBED'
+            agegate_ytcfg['INNERTUBE_CONTEXT']['thirdParty'] = third_party
             agegate_ytcfg['priority'] -= 1
         elif client.endswith('_embedded'):
+            ytcfg['INNERTUBE_CONTEXT']['thirdParty'] = third_party
             ytcfg['priority'] -= 2
         else:
             ytcfg['priority'] -= 3
@@ -200,23 +235,28 @@ def build_innertube_clients():
 
 class YoutubeBaseInfoExtractor(InfoExtractor):
     """Provide base functions for Youtube extractors"""
-    _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
-    _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
-
-    _LOOKUP_URL = 'https://accounts.google.com/_/signin/sl/lookup'
-    _CHALLENGE_URL = 'https://accounts.google.com/_/signin/sl/challenge'
-    _TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}'
 
     _RESERVED_NAMES = (
-        r'channel|c|user|browse|playlist|watch|w|v|embed|e|watch_popup|shorts|'
-        r'movies|results|shared|hashtag|trending|feed|feeds|oembed|get_video_info|'
+        r'channel|c|user|playlist|watch|w|v|embed|e|watch_popup|'
+        r'shorts|movies|results|shared|hashtag|trending|feed|feeds|'
+        r'browse|oembed|get_video_info|iframe_api|s/player|'
         r'storefront|oops|index|account|reporthistory|t/terms|about|upload|signin|logout')
 
+    _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM|WL|LL|LM)'
+
     _NETRC_MACHINE = 'youtube'
+
     # If True it will raise an error if no login info is provided
     _LOGIN_REQUIRED = False
 
-    _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM|WL|LL|LM)'
+    r'''  # Unused since login is broken
+    _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
+    _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
+
+    _LOOKUP_URL = 'https://accounts.google.com/_/signin/sl/lookup'
+    _CHALLENGE_URL = 'https://accounts.google.com/_/signin/sl/challenge'
+    _TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}'
+    '''
 
     def _login(self):
         """
@@ -455,10 +495,9 @@ def _ytcfg_get_safe(self, ytcfg, getter, expected_type=None, default_client='web
         return _func(ytcfg) or _func(self._get_default_ytcfg(default_client))
 
     def _extract_client_name(self, ytcfg, default_client='web'):
-        return (
-            try_get(ytcfg, lambda x: x['INNERTUBE_CLIENT_NAME'], compat_str)
-            or self._ytcfg_get_safe(
-                ytcfg, lambda x: x['INNERTUBE_CONTEXT']['client']['clientName'], compat_str, default_client))
+        return self._ytcfg_get_safe(
+            ytcfg, (lambda x: x['INNERTUBE_CLIENT_NAME'],
+                    lambda x: x['INNERTUBE_CONTEXT']['client']['clientName']), compat_str, default_client)
 
     @staticmethod
     def _extract_session_index(*data):
@@ -468,10 +507,9 @@ def _extract_session_index(*data):
                 return session_index
 
     def _extract_client_version(self, ytcfg, default_client='web'):
-        return (
-            try_get(ytcfg, lambda x: x['INNERTUBE_CLIENT_VERSION'], compat_str)
-            or self._ytcfg_get_safe(
-                ytcfg, lambda x: x['INNERTUBE_CONTEXT']['client']['clientVersion'], compat_str, default_client))
+        return self._ytcfg_get_safe(
+            ytcfg, (lambda x: x['INNERTUBE_CLIENT_VERSION'],
+                    lambda x: x['INNERTUBE_CONTEXT']['client']['clientVersion']), compat_str, default_client)
 
     def _extract_api_key(self, ytcfg=None, default_client='web'):
         return self._ytcfg_get_safe(ytcfg, lambda x: x['INNERTUBE_API_KEY'], compat_str, default_client)
@@ -496,24 +534,31 @@ def _extract_context(self, ytcfg=None, default_client='web'):
             context['client']['visitorData'] = visitor_data
         return context
 
+    _SAPISID = None
+
     def _generate_sapisidhash_header(self, origin='https://www.youtube.com'):
-        # Sometimes SAPISID cookie isn't present but __Secure-3PAPISID is.
-        # See: https://github.com/yt-dlp/yt-dlp/issues/393
-        yt_cookies = self._get_cookies('https://www.youtube.com')
-        sapisid_cookie = dict_get(
-            yt_cookies, ('__Secure-3PAPISID', 'SAPISID'))
-        if sapisid_cookie is None or not sapisid_cookie.value:
-            return
         time_now = round(time.time())
-        # SAPISID cookie is required if not already present
-        if not yt_cookies.get('SAPISID'):
-            self.write_debug('Copying __Secure-3PAPISID cookie to SAPISID cookie', only_once=True)
-            self._set_cookie(
-                '.youtube.com', 'SAPISID', sapisid_cookie.value, secure=True, expire_time=time_now + 3600)
-        self.write_debug('Extracted SAPISID cookie', only_once=True)
+        if self._SAPISID is None:
+            yt_cookies = self._get_cookies('https://www.youtube.com')
+            # Sometimes SAPISID cookie isn't present but __Secure-3PAPISID is.
+            # See: https://github.com/yt-dlp/yt-dlp/issues/393
+            sapisid_cookie = dict_get(
+                yt_cookies, ('__Secure-3PAPISID', 'SAPISID'))
+            if sapisid_cookie and sapisid_cookie.value:
+                self._SAPISID = sapisid_cookie.value
+                self.write_debug('Extracted SAPISID cookie')
+                # SAPISID cookie is required if not already present
+                if not yt_cookies.get('SAPISID'):
+                    self.write_debug('Copying __Secure-3PAPISID cookie to SAPISID cookie')
+                    self._set_cookie(
+                        '.youtube.com', 'SAPISID', self._SAPISID, secure=True, expire_time=time_now + 3600)
+            else:
+                self._SAPISID = False
+        if not self._SAPISID:
+            return None
         # SAPISIDHASH algorithm from https://stackoverflow.com/a/32065323
         sapisidhash = hashlib.sha1(
-            f'{time_now} {sapisid_cookie.value} {origin}'.encode('utf-8')).hexdigest()
+            f'{time_now} {self._SAPISID} {origin}'.encode('utf-8')).hexdigest()
         return f'SAPISIDHASH {time_now}_{sapisidhash}'
 
     def _call_api(self, ep, query, video_id, fatal=True, headers=None,
@@ -676,11 +721,11 @@ def _extract_alerts(cls, data):
                 if message:
                     yield alert_type, message
 
-    def _report_alerts(self, alerts, expected=True):
+    def _report_alerts(self, alerts, expected=True, fatal=True):
         errors = []
         warnings = []
         for alert_type, alert_message in alerts:
-            if alert_type.lower() == 'error':
+            if alert_type.lower() == 'error' and fatal:
                 errors.append([alert_type, alert_message])
             else:
                 warnings.append([alert_type, alert_message])
@@ -746,6 +791,13 @@ def _extract_response(self, item_id, query, note='Downloading API JSON', headers
                     note='%s%s' % (note, ' (retry #%d)' % count if count else ''))
             except ExtractorError as e:
                 if isinstance(e.cause, network_exceptions):
+                    if isinstance(e.cause, compat_HTTPError) and not is_html(e.cause.read(512)):
+                        e.cause.seek(0)
+                        yt_error = try_get(
+                            self._parse_json(e.cause.read().decode(), item_id, fatal=False),
+                            lambda x: x['error']['message'], compat_str)
+                        if yt_error:
+                            self._report_alerts([('ERROR', yt_error)], fatal=False)
                     # Downloading page may result in intermittent 5xx HTTP error
                     # Sometimes a 404 is also recieved. See: https://github.com/ytdl-org/youtube-dl/issues/28289
                     # We also want to catch all other network exceptions since errors in later pages can be troublesome
@@ -890,7 +942,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             youtube\.googleapis\.com)/                        # the various hostnames, with wildcard subdomains
                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
                          (?:                                                  # the various things that can precede the ID:
-                             (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
+                             (?:(?:v|embed|e|shorts)/(?!videoseries))         # v/ or embed/ or e/ or shorts/
                              |(?:                                             # or the v= param in all its forms
                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
@@ -1017,24 +1069,17 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         '_rtmp': {'protocol': 'rtmp'},
 
         # av01 video only formats sometimes served with "unknown" codecs
-        '394': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
-        '395': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
-        '396': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
-        '397': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
+        '394': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'vcodec': 'av01.0.00M.08'},
+        '395': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'av01.0.00M.08'},
+        '396': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'av01.0.01M.08'},
+        '397': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'av01.0.04M.08'},
+        '398': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'av01.0.05M.08'},
+        '399': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'av01.0.08M.08'},
+        '400': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'av01.0.12M.08'},
+        '401': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'av01.0.12M.08'},
     }
     _SUBTITLE_FORMATS = ('json3', 'srv1', 'srv2', 'srv3', 'ttml', 'vtt')
 
-    _AGE_GATE_REASONS = (
-        'Sign in to confirm your age',
-        'This video may be inappropriate for some users.',
-        'Sorry, this content is age-restricted.',
-        'Please confirm your age.')
-
-    _AGE_GATE_STATUS_REASONS = (
-        'AGE_VERIFICATION_REQUIRED',
-        'AGE_CHECK_REQUIRED'
-    )
-
     _GEO_BYPASS = False
 
     IE_NAME = 'youtube'
@@ -1139,8 +1184,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'format': '141/bestaudio[ext=m4a]',
             },
         },
-        # Normal age-gate video (embed allowed)
+        # Age-gate videos. See https://github.com/yt-dlp/yt-dlp/pull/575#issuecomment-888837000
         {
+            'note': 'Embed allowed age-gate video',
             'url': 'https://youtube.com/watch?v=HtVdAasjOgU',
             'info_dict': {
                 'id': 'HtVdAasjOgU',
@@ -1155,6 +1201,52 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'age_limit': 18,
             },
         },
+        {
+            'note': 'Age-gate video with embed allowed in public site',
+            'url': 'https://youtube.com/watch?v=HsUATh_Nc2U',
+            'info_dict': {
+                'id': 'HsUATh_Nc2U',
+                'ext': 'mp4',
+                'title': 'Godzilla 2 (Official Video)',
+                'description': 'md5:bf77e03fcae5529475e500129b05668a',
+                'upload_date': '20200408',
+                'uploader_id': 'FlyingKitty900',
+                'uploader': 'FlyingKitty',
+                'age_limit': 18,
+            },
+        },
+        {
+            'note': 'Age-gate video embedable only with clientScreen=EMBED',
+            'url': 'https://youtube.com/watch?v=Tq92D6wQ1mg',
+            'info_dict': {
+                'id': 'Tq92D6wQ1mg',
+                'title': '[MMD] Adios - EVERGLOW [+Motion DL]',
+                'ext': 'mp4',
+                'upload_date': '20191227',
+                'uploader_id': 'UC1yoRdFoFJaCY-AGfD9W0wQ',
+                'uploader': 'Projekt Melody',
+                'description': 'md5:17eccca93a786d51bc67646756894066',
+                'age_limit': 18,
+            },
+        },
+        {
+            'note': 'Non-Agegated non-embeddable video',
+            'url': 'https://youtube.com/watch?v=MeJVWBSsPAY',
+            'info_dict': {
+                'id': 'MeJVWBSsPAY',
+                'ext': 'mp4',
+                'title': 'OOMPH! - Such Mich Find Mich (Lyrics)',
+                'uploader': 'Herr Lurik',
+                'uploader_id': 'st3in234',
+                'description': 'Fan Video. Music & Lyrics by OOMPH!.',
+                'upload_date': '20130730',
+            },
+        },
+        {
+            'note': 'Non-bypassable age-gated video',
+            'url': 'https://youtube.com/watch?v=Cr381pDsSsA',
+            'only_matching': True,
+        },
         # video_info is None (https://github.com/ytdl-org/youtube-dl/issues/4421)
         # YouTube Red ad is not captured for creator
         {
@@ -1323,6 +1415,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             'params': {
                 'skip_download': True,
             },
+            'skip': 'Not multifeed anymore',
         },
         {
             # Multifeed video with comma in title (see https://github.com/ytdl-org/youtube-dl/issues/8536)
@@ -1739,14 +1832,17 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             'params': {
                 'extractor_args': {'youtube': {'player_skip': ['configs']}},
             },
-        }
+        }, {
+            # shorts
+            'url': 'https://www.youtube.com/shorts/BGQWPY4IigY',
+            'only_matching': True,
+        },
     ]
 
     @classmethod
     def suitable(cls, url):
-        # Hack for lazy extractors until more generic solution is implemented
-        # (see #28780)
-        from .youtube import parse_qs
+        from ..utils import parse_qs
+
         qs = parse_qs(url)
         if qs.get('list', [None])[0]:
             return False
@@ -1861,10 +1957,10 @@ def _parse_sig_js(self, jscode):
         funcname = self._search_regex(
             (r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
-             r'\bm=(?P<sig>[a-zA-Z0-9$]{2})\(decodeURIComponent\(h\.s\)\)',
-             r'\bc&&\(c=(?P<sig>[a-zA-Z0-9$]{2})\(decodeURIComponent\(c\)\)',
-             r'(?:\b|[^a-zA-Z0-9$])(?P<sig>[a-zA-Z0-9$]{2})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\);[a-zA-Z0-9$]{2}\.[a-zA-Z0-9$]{2}\(a,\d+\)',
-             r'(?:\b|[^a-zA-Z0-9$])(?P<sig>[a-zA-Z0-9$]{2})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
+             r'\bm=(?P<sig>[a-zA-Z0-9$]{2,})\(decodeURIComponent\(h\.s\)\)',
+             r'\bc&&\(c=(?P<sig>[a-zA-Z0-9$]{2,})\(decodeURIComponent\(c\)\)',
+             r'(?:\b|[^a-zA-Z0-9$])(?P<sig>[a-zA-Z0-9$]{2,})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\);[a-zA-Z0-9$]{2}\.[a-zA-Z0-9$]{2}\(a,\d+\)',
+             r'(?:\b|[^a-zA-Z0-9$])(?P<sig>[a-zA-Z0-9$]{2,})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
              r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
              # Obsolete patterns
              r'(["\'])signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
@@ -1996,8 +2092,7 @@ def extract_id(cls, url):
         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
         if mobj is None:
             raise ExtractorError('Invalid URL: %s' % url)
-        video_id = mobj.group(2)
-        return video_id
+        return mobj.group('id')
 
     def _extract_chapters_from_json(self, data, duration):
         chapter_list = traverse_obj(
@@ -2342,14 +2437,21 @@ def _generate_player_context(sts=None):
             'racyCheckOk': True
         }
 
-    def _is_agegated(self, player_response):
-        reasons = traverse_obj(player_response, ('playabilityStatus', ('status', 'reason')), default=[])
-        for reason in reasons:
-            if reason in self._AGE_GATE_REASONS + self._AGE_GATE_STATUS_REASONS:
-                return True
-        if traverse_obj(player_response, ('playabilityStatus', 'desktopLegacyAgeGateReason')) is not None:
+    @staticmethod
+    def _is_agegated(player_response):
+        if traverse_obj(player_response, ('playabilityStatus', 'desktopLegacyAgeGateReason')):
             return True
-        return False
+
+        reasons = traverse_obj(player_response, ('playabilityStatus', ('status', 'reason')), default=[])
+        AGE_GATE_REASONS = (
+            'confirm your age', 'age-restricted', 'inappropriate',  # reason
+            'age_verification_required', 'age_check_required',  # status
+        )
+        return any(expected in reason for expected in AGE_GATE_REASONS for reason in reasons)
+
+    @staticmethod
+    def _is_unplayable(player_response):
+        return traverse_obj(player_response, ('playabilityStatus', 'status')) == 'UNPLAYABLE'
 
     def _extract_player_response(self, client, video_id, master_ytcfg, player_ytcfg, identity_token, player_url, initial_pr):
 
@@ -2364,7 +2466,7 @@ def _extract_player_response(self, client, video_id, master_ytcfg, player_ytcfg,
         yt_query.update(self._generate_player_context(sts))
         return self._extract_response(
             item_id=video_id, ep='player', query=yt_query,
-            ytcfg=player_ytcfg, headers=headers, fatal=False,
+            ytcfg=player_ytcfg, headers=headers, fatal=True,
             default_client=client,
             note='Downloading %s player API JSON' % client.replace('_', ' ').strip()
         ) or None
@@ -2386,7 +2488,7 @@ def _get_requested_clients(self, url, smuggled_data):
 
         if smuggled_data.get('is_music_url') or self.is_music_url(url):
             requested_clients.extend(
-                f'{client}_music' for client in requested_clients if not client.endswith('_music'))
+                f'{client}_music' for client in requested_clients if f'{client}_music' in INNERTUBE_CLIENTS)
 
         return orderedSet(requested_clients)
 
@@ -2409,31 +2511,52 @@ def _extract_player_responses(self, clients, video_id, webpage, master_ytcfg, pl
 
         original_clients = clients
         clients = clients[::-1]
+
+        def append_client(client_name):
+            if client_name in INNERTUBE_CLIENTS and client_name not in original_clients:
+                clients.append(client_name)
+
+        # Android player_response does not have microFormats which are needed for
+        # extraction of some data. So we return the initial_pr with formats
+        # stripped out even if not requested by the user
+        # See: https://github.com/yt-dlp/yt-dlp/issues/501
+        yielded_pr = False
+        if initial_pr:
+            pr = dict(initial_pr)
+            pr['streamingData'] = None
+            yielded_pr = True
+            yield pr
+
+        last_error = None
         while clients:
             client = clients.pop()
             player_ytcfg = master_ytcfg if client == 'web' else {}
             if 'configs' not in self._configuration_arg('player_skip'):
                 player_ytcfg = self._extract_player_ytcfg(client, video_id) or player_ytcfg
 
-            pr = (
-                initial_pr if client == 'web' and initial_pr
-                else self._extract_player_response(
-                    client, video_id, player_ytcfg or master_ytcfg, player_ytcfg, identity_token, player_url, initial_pr))
+            try:
+                pr = initial_pr if client == 'web' and initial_pr else self._extract_player_response(
+                    client, video_id, player_ytcfg or master_ytcfg, player_ytcfg, identity_token, player_url, initial_pr)
+            except ExtractorError as e:
+                if last_error:
+                    self.report_warning(last_error)
+                last_error = e
+                continue
+
             if pr:
+                yielded_pr = True
                 yield pr
 
-            if self._is_agegated(pr):
-                client = f'{client}_agegate'
-                if client in INNERTUBE_CLIENTS and client not in original_clients:
-                    clients.append(client)
+            # creator clients can bypass AGE_VERIFICATION_REQUIRED if logged in
+            if client.endswith('_agegate') and self._is_unplayable(pr) and self._generate_sapisidhash_header():
+                append_client(client.replace('_agegate', '_creator'))
+            elif self._is_agegated(pr):
+                append_client(f'{client}_agegate')
 
-        # Android player_response does not have microFormats which are needed for
-        # extraction of some data. So we return the initial_pr with formats
-        # stripped out even if not requested by the user
-        # See: https://github.com/yt-dlp/yt-dlp/issues/501
-        if initial_pr and 'web' not in original_clients:
-            initial_pr['streamingData'] = None
-            yield initial_pr
+        if last_error:
+            if not yielded_pr:
+                raise last_error
+            self.report_warning(last_error)
 
     def _extract_formats(self, streaming_data, video_id, player_url, is_live):
         itags, stream_ids = [], []
@@ -2531,7 +2654,9 @@ def _extract_formats(self, streaming_data, video_id, player_url, is_live):
             yield dct
 
         skip_manifests = self._configuration_arg('skip')
-        get_dash = not is_live and 'dash' not in skip_manifests and self.get_param('youtube_include_dash_manifest', True)
+        get_dash = (
+            (not is_live or self._configuration_arg('include_live_dash'))
+            and 'dash' not in skip_manifests and self.get_param('youtube_include_dash_manifest', True))
         get_hls = 'hls' not in skip_manifests and self.get_param('youtube_include_hls_manifest', True)
 
         def guess_quality(f):
@@ -2668,8 +2793,7 @@ def feed_entry(name):
 
         if not formats:
             if not self.get_param('allow_unplayable_formats') and traverse_obj(streaming_data, (..., 'licenseInfos')):
-                self.raise_no_formats(
-                    'This video is DRM protected.', expected=True)
+                self.report_drm(video_id)
             pemr = get_first(
                 playability_statuses,
                 ('errorScreen', 'playerErrorMessageRenderer'), expected_type=dict) or {}
@@ -2689,8 +2813,8 @@ def feed_entry(name):
         for f in formats:
             if '&c=WEB&' in f['url'] and '&ratebypass=yes&' not in f['url']:  # throttled
                 f['source_preference'] = -10
-                note = f.get('format_note')
-                f['format_note'] = f'{note} (throttled)' if note else '(throttled)'
+                # TODO: this method is not reliable
+                f['format_note'] = format_field(f, 'format_note', '%s ') + '(maybe throttled)'
 
         # Source is given priority since formats that throttle are given lower source_preference
         # When throttling issue is fully fixed, remove this
@@ -3361,7 +3485,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
     }, {
         'url': 'https://www.youtube.com/channel/UCoMdktPbSTixAyNGwb-UYkQ/live',
         'info_dict': {
-            'id': 'FMtPN8yp5LU',  # This will keep changing
+            'id': '3yImotZU3tw',  # This will keep changing
             'ext': 'mp4',
             'title': compat_str,
             'uploader': 'Sky News',
@@ -4330,7 +4454,7 @@ class YoutubeYtBeIE(InfoExtractor):
     }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+        mobj = self._match_valid_url(url)
         video_id = mobj.group('id')
         playlist_id = mobj.group('playlist_id')
         return self.url_result(
@@ -4473,7 +4597,7 @@ def _make_valid_url(cls):
         return cls._VALID_URL
 
     def _real_extract(self, url):
-        qs = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
+        qs = parse_qs(url)
         query = (qs.get('search_query') or qs.get('q'))[0]
         self._SEARCH_PARAMS = qs.get('sp', ('',))[0]
         return self._get_n_results(query, self._MAX_RESULTS)